3 ms·
I was going to say that this is what we used as one of two entropy sources [0] in the Cryptech HSM. But the second link there is the fork of the Cryptech rosc e
by JoachimS 4y ago
I was going to say that this is what we used as one of two entropy sources [0] in the Cryptech HSM. But the second link there is the fork of the Cryptech rosc entropy source we did for the FPGA based NTS server. ;-)
The TRNG in the Tillitis Key 1 (TK1) we released a couple of weeks ago contain a version of the rosc entropy source. But I have now included a VN-decorrelator. And the actual oscillators are much simpler, cleaner.
[0] The other entropy source in cryptech is an avalanche noise source outside of the FPGA. Both sources can be found here:
https://git.cryptech.is/core/rng/avalanche_entropy/ https://git.cryptech.is/core/rng/avalanche_entropy/
https://git.cryptech.is/core/rng/rosc_entropy/ https://git.cryptech.is/core/rng/rosc_entropy/
The complete TRNG samples both sources (round robin) and creates a seed using SHA-512, with previous seed a part of the input. The seed is then used to initialize ChaCha20 as CSPRNG:
https://git.cryptech.is/core/rng/trng/ https://git.cryptech.is/core/rng/trng/
- ur-whale 4y ago> But I have now included a VN-decorrelator Yeah, but Von-Neuman decorrelators have one major drawback: their throughput is unpredictable.